Platform-agnostic predictive models based on database management system instructions
Platform-agnostic predictive models based on database management system instructions are described. A system identifies a representation of data transformations associated with a first predictive model that executes on a first computing platform. The system parses the representation of data transformations. The system generates database management system instructions that correspond to the parsed representation of data transformations. The system sends the database management system instructions to a second predictive model that executes on a second computing platform, thereby enabling the second predictive model to execute at least some of the database management system instructions to generate a prediction. The first computing platform and the second computing platform are different types of computing platforms.
1. A system comprising:
one or more processors; and
a non-transitory computer readable medium storing a plurality of instructions, which when executed, cause the one or more processors to:
identify a representation of data transformations associated with a first predictive model that executes on a first computing platform, wherein the representation defines sequences of transformations performed on data before, during, and after application of the first predictive model;
parse the representation of data transformations;
generate database management system instructions that correspond to the parsed representation of data transformations; and
send the database management system instructions to a second predictive model that executes on a second computing platform, thereby enabling the second predictive model to execute at least some of the database management system instructions to generate a prediction, wherein the first computing platform and the second computing platform are different types of computing platforms;
wherein the parsed representation of data transformations includes pre-processing data transformations, the first predictive model's data transformation, and post-processing data transformations that can be utilized by the second predictive model to replay the sequence of transformations needed to transform data;
wherein the database management system instructions enable platform-agnostic predictive models;
wherein parsing the representation of data transformations comprises modifying a predictive analytics framework to emit database management system instructions that correspond to the representation of data transformations in response to receiving requests for functions that correspond to the representation of data transformations.
2. The system of claim 1 , wherein parsing the representation of data transformations comprises parsing the representation of data transformations when the first predictive model lacks a current execution status.
3. The system of claim 1 , wherein parsing the representation of data transformations comprises dynamically recording database management system instructions that are generated in response to the first predictive model implementing the representation of data transformations.
4. The system of claim 1 , wherein generating the database management system instructions that correspond to the parsed representation of data transformations comprises deleting some of the database management system instructions to optimize the database management system instructions, wherein some of the database management system instructions deleted include instructions associated with a structured query language.
5. The system of claim 1 , wherein generating the database management system instructions that correspond to the parsed representation of data transformations comprises identifying data upon which other data depends, and injecting the data upon which the other data depends into the database management system instructions.
6. The system of claim 1 , wherein the database management system instructions that correspond to the parsed representation of data transformations comprises code associated with a structured query language.
7. A method comprising:
identifying a representation of data transformations associated with a first predictive model that executes on a first computing platform, wherein the representation defines sequences of transformations performed on data before, during, and after application of the first predictive model;
parsing the representation of data transformations;
generating database management system instructions that correspond to the parsed representation of data transformations; and
sending the database management system instructions to a second predictive model that executes on a second computing platform, thereby enabling the second predictive model to execute at least some of the database management system instructions to generate a prediction, wherein the first computing platform and the second computing platform are different types of computing platforms;
wherein the parsed representation of data transformations includes pre-processing data transformations, the first predictive model's data transformation, and post-processing data transformations that can be utilized by the second predictive model to replay the sequence of transformations needed to transform data;
wherein the database management system instructions enable platform-agnostic predictive models;
wherein parsing the representation of data transformations comprises modifying a predictive analytics framework to emit database management system instructions that correspond to the representation of data transformations in response to receiving requests for functions that correspond to the representation of data transformations.
8. The computer-implemented method of claim 7 , wherein parsing the representation of data transformations comprises parsing the representation of data transformations when the first predictive model lacks a current execution status.
9. The computer-implemented method of claim 7 , wherein parsing the representation of data transformations comprises dynamically recording database management system instructions that are generated in response to the first predictive model implementing the representation of data transformations.
10. The computer-implemented method of claim 7 , wherein generating the database management system instructions that correspond to the parsed representation of data transformations comprises deleting some of the database management system instructions to optimize the database management system instructions, wherein some of the database management system instructions deleted include instructions associated with a structured query language.
11. The computer-implemented method of claim 7 , wherein generating the database management system instructions that correspond to the parsed representation of data transformations comprises identifying data upon which other data depends, and injecting the data upon which the other data depends into the database management system instructions.
12. The computer-implemented method of claim 7 , wherein the database management system instructions that correspond to the parsed representation of data transformations comprises code associated with a structured query language.
13. A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein to be executed by one or more processors, the program code including instructions to:
identify a representation of data transformations associated with a first predictive model that executes on a first computing platform, wherein the representation defines sequences of transformations performed on data before, during, and after application of the first predictive model;
parse the representation of data transformations;
generate database management system instructions that correspond to the parsed representation of data transformations; and
send the database management system instructions to a second predictive model that executes on a second computing platform, thereby enabling the second predictive model to execute at least some of the database management system instructions to generate a prediction, wherein the first computing platform and the second computing platform are different types of computing platforms;
wherein the parsed representation of data transformations includes pre-processing data transformations, the first predictive model's data transformation, and post-processing data transformations that can be utilized by the second predictive model to replay the sequence of transformations needed to transform data;
wherein the database management system instructions enable platform-agnostic predictive models;
wherein parsing the representation of data transformations comprises modifying a predictive analytics framework to emit database management system instructions that correspond to the representation of data transformations in response to receiving requests for functions that correspond to the representation of data transformations.
14. The computer program product of claim 13 , wherein parsing the representation of data transformations comprises parsing the representation of data transformations when the first predictive model lacks a current execution status.
15. The computer program product of claim 13 , wherein parsing the representation of data transformations comprises dynamically recording database management system instructions that are generated in response to the first predictive model implementing the representation of data transformations.
16. The computer program product of claim 13 , wherein generating the database management system instructions that correspond to the parsed representation of data transformations comprises deleting some of the database management system instructions to optimize the database management system instructions, wherein some of the database management system instructions deleted include instructions associated with a structured query language.
17. The computer program product of claim 13 , wherein generating the database management system instructions that correspond to the parsed representation of data transformations comprises identifying data upon which other data depends, and injecting the data upon which the other data depends into the database management system instructions, and the database management system instructions that correspond to the parsed representation of data transformations comprises code associated with a structured query language.